At the end of the _WM_Print.log file is a summary of the simulation effort for each of the MSHE processes.
The best way to speed up the simulation is to find which process is taking the most time and work on it first.
In the screen shot that you included, all of your timesteps are 24 hours.
This is much too long for any of the very dynamic processes, such as OL, UZ and the river.
So, I suspect that the simulation is simply not converging.
If you really don't need short time steps, then I suggest that you use the 2LayerUZ and routing branches in the rivers.
If you are not worried about the OL dynamics, you can also use a low ManningsM (very rough) to slow the OL velocities.
Probably the most important thing is to start out with short simulations and a coarse grid. When you have a stable simulation, you can start slowly lengthening the simulation period and then refine the grid.
Regards,
Douglas
